JOURNAL OF COMPUTERS, VOL. 4, NO.

8, AUGUST 2009 763

Texture Image Classification Using Visual

Perceptual Texture Features and Gabor Wavelet
Features
Muwei Jian
School of Space Science and Physics, Shandong University at Weihai, Weihai, China
Email: jianmuwei@gmail.com

Haoyan GUO
School of Computer Science and Technology, Harbin Institute of Technology (Weihai), Weihai, China

Lei Liu
Institute of Physics and Electronic Engineering, TaiShan University, Taian, China

Abstract—Texture can describe a wide variety of surface
characteristics and a key component for human visual
perception and plays an important role in image-related
applications. This paper proposes a scheme for texture
image classification using visual perceptual texture features
and Gabor wavelet features. Three new texture features
which are proved to be in accordance with human visual
perceptions are introduced. Usually, Subband statistics
based on Gabor wavelet features are normally used to
construct feature vectors for texture image classification.
However, most previous methods make no further analysis
of the decomposed subbands or simply remove most detail
coefficients. The classification algorithms commonly use
many features without consideration of whether the features
are effective for discriminating different classes. This may
produce unnecessary computation burden and even
decrease the retrieval performance. This paper proposes a
method for selecting effective Gabor wavelet subbands
based on feature selection functions. The method can
discard those subbands that are redundant or may lead to
wrong classification results. We test our proposed method
using the Brodatz texture database, and the experimental
results show the scheme has produced promising results.
Index Terms—Visual Perception Texture Features, Gabor
wavelet features; Texture image classification; SVM
I. INTRODUCTION
Texture is an important component of human visual
perception and can be effectively used for characterizing
different image regions. For example, the skin of a zebra
can be seen as texture with black and white stripe pattern.
In the past, texture features have been extensively studied
in the research area of texture image classification and
content-based image retrieval, as well as other fields
related to pattern analysis [1, 2, 3, 4].
Basically, texture representation methods can be
classified into three categories, namely structural,
statistical and multi-resolution filtering methods. Typical
structure-based methods include morphological and graph
techniques, which describe texture using structural
primitives and layout [1]. Statistical methods are
commonly used and proved to be effective in texture
analysis [3, 4, 5]. Methods based on Multi-resolution
decompose a texture image into different scales, from
which more statistics can be extracted and used to
describe texture features [7, 8, 11, 12, 13]. These methods
have been effectively used for solving texture recognition
problems.
Due to the resemblance between multi-resolution
filtering techniques (such as Gabor and wavelet
transform) and human visual process, Gabor and wavelet
transform techniques are often used for texture
characterization through the analysis of spatial-frequency
content. Many publications showed the effectiveness of
using these techniques for texture analysis, segmentation,
retrieval and classification [7, 8, 11, 12, 13].
However, most previous methods make no further
analysis of the decomposed subbands or simply remove
most detail coefficients. The classification algorithms
commonly use many features without consideration of
whether the features are effective for discriminating
different classes. This may result in unnecessary
computation and even decrease the classification
performance.
Psychophysical investigation has shown that the Human
Visual System (HVS) does a frequency analysis when we
see images [22, 23, 24]. Texture is especially suited for
this type of analysis for its intrinsic properties. In this
paper, we introduce three visual perceptual features,
namely directionality, contrast and coarseness, which
have proved to be in accordance with human visual
perception in image retrieval experiments [17, 18].
Then, a scheme for texture image classification using
visual perceptual texture features and Gabor wavelet
features is proposed. In contrast to previous work, our
study focuses on reducing the dimensionality of feature
vectors by discarding those subbands that may be
irrelevant or redundant, with the aim of achieving better

© 2009 ACADEMY PUBLISHER

764
texture classification performance. We introduce a
procedure for selecting effective wavelet subbands [16].
The feature selection functions are used to rank all
subbands, so that the problem of selecting the most
“useful” subband set reduces to picking up those with
highest rankings. Statistics extracted from such subbands
are then used to construct feature vectors. In this way, the
dimensionality of feature vectors is reduced. Experiments
show the selected subbands are effective for
discriminating different textures, as the classification
performance is much improved.
The rest of the paper is organized as follows. In section
2, we introduce the visual perceptual texture features.
Section 3 briefly describes Gabor wavelet decomposition.
In section 4, a feature selection algorithm is introduced.
In section 5, experimental results are presented in section
5. Finally, we conclude the paper in section 6.
II. NEW TEXTURE FEATURES DERIVING
WAVELET TRANSFORM
Wavelet transform has been widely used in image
processing and shows tremendous advantages over
Fourier transform. Wavelet transform is a multi-
resolution analysis that represents image variations at
different scales [6, 7, 8]. According to the definition of
wavelet, a wavelet is an oscillating and attenuated
function and its integrals equal to zero. It is a
mathematical function useful in digital signal and image
processing.
Given f(x) is a one-dimensional input signal, a 1-
Ddiscrete wavelet transform is defined as:
φ jk ( x) = 2 − j / 2 φ (2 − j x − k ) ,
− j/2
ψ jk ( x ) = 2 ψ (2 − j x − k ) ,
Where: φ (x) and ψ (x ) are the scaling function and
wavelet function respectively,
{φ jk ( x)} and {ψ jk ( x)} are the two orthogonal function
basis sets.
Define P0f=f, a 1-D discrete wavelet transform at the
scale j decomposed Pj-1f through orthogonal projection Pjf
and Qjf as follows:
Pj −1 f = Pj f + Q j f = ∑ c kj φ jk + ∑ d kjψ jk ;
k k
p −1
and c kj = ∑ h(n)c 2jk−1+ n ,
n =0
p −1
d kj = ∑ g (n)c 2jk−+1 n ,
n =0
( j = 1,2,..., L, k = 0,1,..., N 2 j − 1) .
© 2009 ACADEMY PUBLISHER
JOURNAL OF COMPUTERS, VOL. 4, NO. 8, AUGUST 2009
where: {h(n)} and {g(n)} are low pass filter and high pass
{C 0 }
filter, respectively; n is the input signal; N is the
length of the input signal; L is the necessary progression.
Wavelet is commonly used for multi-resolution analysis
and representing image variations at different scales. The
computation framework of the wavelet transform of a 2D
signal involves recursive filtering and sub-sampling, as
shown in Fig.1. To clarify the use of symbols, we briefly
describe the general process of wavelet transforms and
the notations of subbands of wavelet decomposition. At
each level, there are three detail images. Following [7],
we denote the detail images (subband) as LH (contains
the high frequency horizontal information), HL (contains
the high frequency vertical information), and HH
(contains the high frequency diagonal information). The
decomposition/transform also produces one
FROM approximation image, denoted by LL, containing the low
frequency information. The wavelet transform can
recursively decompose the LL band. Since 2 level
wavelet decomposition yields 6 detail images, we use
LH1, HL1, HH1, LH2, HL2, HH2, and an additional
approximation image LL2 to denote all the subband
images.
LL2 HL2
HL1
LH2 HH2
LH1 HH1
Fig.1. Two-level wavelet decomposition
A. Directionality
Directionality is an important characteristic for texture
images. For example, from human perception’s viewpoint,
we perceive the D11 texture in Figure 2 as a “vertical”
texture, D49 as a “horizontal” one and D47as “diagonal”.
Instead of computing a vague value of directionality, we
introduce three different directionalities, namely “the
vertical directionality”, “the horizontal directionality” and
“the diagonal directionality”, to represent the directional
information of a texture image.
JOURNAL OF COMPUTERS, VOL. 4, NO. 8, AUGUST 2009 765

200

100

D11 D49 D47

-100

-200
Fig.2 Three textures from Brodatz database.
-300

The Horizontal Directionality

-400

In allusion to every HL (contains the high vertical 0 20 40 60 80 100 120 140

frequency information) subband of wavelet (b)

decomposition, we compute the horizontal directionality.
Let M, N be the sizes of HL subband, and x ( j , k ) be the Fig.3 (a) The D1 texture from Brodatz database. (b). The coefficient
subband’s coefficient of wavelet decomposition, where j convolution result of the 10th row’s HL1 subband .
and k represent the row and column values of the subband
images respectively. Firstly, the subband image is The Vertical Directionality
⎛ 1 0 1⎞ Similar to the definition of the horizontal directionality,
⎜ ⎟ the LH subband is convoluted with a different template
convoluted with a template ⎜ 1 0 1⎟ at each wavelet
⎜ 1 0 1⎟ ⎛1 1 1 ⎞
⎝ ⎠ ⎜ ⎟
decomposition coefficient in order to enhance the ⎜ 0 0 0 ⎟ for each coefficient produced by wavelet
direction contrast. Let h ( j , k ) represent the result of ⎜1 1 1 ⎟
⎝ ⎠
convolution. For every row, the normalized convolution decomposition. Let v ( j , k ) represent the result of
result is computed as: convolution; j and k represent the row and column values
of the subband. Then for every column, normalized
| h( j , k ) | coefficient convolution result is computed as:
p jk = N
;
∑ | h( j , k ) | | v( j , k ) |
k =1
q jk = M
;
The horizontal directionality is then defined as : ∑ | v( j , k ) |
j =1

M N
1
DirH =
MN
∑∑ p
j =1 k =1
jk
2
, The vertical directionality is defined as :

M N
1
Where M, N are the sizes of HL subband.
DirV =
MN
∑∑q
j =1 k =1
jk
2
,
Fig. 3 (a) is the D1 texture from the Brodatz database and
(b) shows the 10th row’s coefficient convolution result of Where M, N are the sizes of HL subband.
HL1 subband .

The Diagonal Directionality

Compared to the horizontal and vertical directionality, the
computation of the diagonal directionality is more
complex. We should take two diagonal directions into
account, namely, π / 4 and 3π / 4 .
When we consider the diagonal directionality of π / 4 ,
⎛0 0 1⎞
(a) ⎜ ⎟
convolution is performed with a template ⎜ 0 1 0 ⎟ at the
⎜1 0 0 ⎟
⎝ ⎠
HH subband. Similarly, when the diagonal directionality
of 3π / 4 is considered, convolution is computed with a
⎛1 0 0 ⎞
⎜ ⎟
template ⎜ 0 1 0 ⎟ . Then we sample π / 4 diagonals and
⎜0 0 1⎟
⎝ ⎠

© 2009 ACADEMY PUBLISHER

766 JOURNAL OF COMPUTERS, VOL. 4, NO. 8, AUGUST 2009

3π / 4 diagonals for computing the diagonal

directionality. The calculation process is the same as the j = s + 2 z −1 k = t + 2 z −1
horizontal and vertical directionality. The diagonal A z ( s, t ) = ∑ ∑ x( j , k )
22 z
directionality in the same wavelet decomposition level is j =s−2 z −1
k =t − 2 z −1
.
defined as the average of the diagonal directionality of
π / 4 and 3π / 4 :
Where x ( j , k ) is the coefficient of the LL subband at
1 ( j, k).
DirD = ( DirD1 + DirD 2 ) , Then, the differences between pairs of non-overlapping
2 moving averages in the horizontal and vertical directions
for each pixel are computed,
Where DirD1 represents the Diagonal Directionality of
π / 4 and DirD 2 represents the Diagonal Directionality z −1 z −1
E z ,h ( s , t ) =| A z ( s + 2 , t ) − Az ( s − 2 , t) |
of 3π / 4 , respectively. ;

B. Contrast z −1 z −1
E z ,v ( s , t ) =| A z ( s , t + 2 ) − Az ( s, t − 2 )|
For the purpose of denoting the change of grey levels .
in a texture, contrast is commonly defined for each pixel
as an estimate of the local variation in a neighborhood. At each point, the value of Z that maximizes E in
The calculation of the contrast is implemented in the either direction is used to set the best size:
wavelet decompose approximation subband, denoted as
LL, which contains the low frequency and reflecting the z
S best ( s , t ) = 2
global information of the texture. ;
Given x ( j , k ) as the coefficient of LL subband, where j
and k represent the row and column values of the The global coarseness is calculated by averaging S best
subband images, respectively, let the pixel at the point (j, over the entire LL subband:

k) be x ( j , k ) and the neighborhood of size W × W be a 1 M N

mask of the pixel. Then local contrast is computed as: Conse =

MN
∑ ∑ S best ( j, k )
j =1 k =1

max x∈W ×W ( x) − min x∈W ×W ( x)

l _ con(i, j ) = More details on the Visual Perception Texture Features
max x∈W ×W ( x) + min x∈W ×W ( x) can be found in [17, 18].
.

The global contrast is defined as the mean of all the local

contrast values: III. GABOR WAVELET FEATURES
Similar to the wavelet transform, Gabor filtering, which is
1 M N also called Gabor wavelet, can also produce a multi-
Con =
MN
∑ ∑ l _ con(i, j )
j =1 k =1
resolution representation of the sample texture image.
. Gabor filtering provides a flexible scheme for designing
efficient algorithms to capture more orientation and scale
Since the wavelet decomposition results in a smaller size information. A two dimensional Gabor function g(x, y) is
for the approximation subband compared with that of the defined as
original texture's size, we recommend a neighborhood
window with a size of 3x3, instead of using the size 7x7 ⎛ ⎛ 2 ⎞
as introduced in [20]. 1 1 x y 2 ⎞⎟
g ( x, y ) = exp ⎜ - ⎜ + + 2π jWx ⎟
2πσ X σ y ⎜⎜ 2 ⎜ 2 σ 2⎟ ⎟⎟
C. Coarseness ⎝ ⎝σX y ⎠ ⎠
Coarseness is the granularity measurement of texture and
the most fundamental texture feature. Researchers usually σ
identify the texture by “ coarseness ” . The coarseness where σ X and y are the standard deviations of the
defined by Tamura etc. in [3] coincides well with the Guassian envelopes along the x and y direction [1].
psychological measurements for human perception, so we Then a set of Gabor filters can be obtained by
also use it in the wavelet decompose approximation appropriate dilations and rotations of g(x, y):
subband to calculate the granularity of texture. The
computational definition of coarseness is briefly g mn(x, y) = a -mg ( x ', y ')
described as follows: ;
A z ( s, t )
The moving average over the
x ' = a −m( x cos θ + y sin θ )
neighborhood of size 2 × 2 (z=0, 1, 2, 3, 4, 5) at the
z z

point (s, t) is y ' = a − m(− x sin θ + y cos θ )

© 2009 ACADEMY PUBLISHER

JOURNAL OF COMPUTERS, VOL. 4, NO. 8, AUGUST 2009 767

(2) Calculate the values of feature selection functions

where a >1, θ = nπ/K, n = 0, 1, …, K-1, and m = 0, 1, …, using the ith-subbands of the query texture and the m
S-1. K and S are the number of orientations and scales. candidate images:
−m
The scale factor a is to ensure that energy is
independent of m. G is1 , … G ij
s ,…, G s ;
im
Given an image I(x, y), its Gabor transform is defined as:
s is the Fisher’s discriminant which is a
where s= L 2 , G ij
∗
W mn = ∫ I ( x, y ) g mn( x − x1, y − y1)dx1dy1
commonly used quantity for feature selection. The basic
idea is to select a feature that maximizes the separation of
Where * indicates the complex conjugate. A bank of
the means of two classes scaled according to the variance.
filters at different scales and orientations can extract the
The quantity for measuring the separation is defined as:
texture frequency and orientation information effectively.
Then the mean µmn and the standard deviation σ mn of ( µ ik − µ jk ) 2
the magnitude of W mn
( x, y )
can be used to construct Gk =
σ ik +σ
2 2
jk
feature vectors, denoted by ; k=1, 2,…,n
f = [ µ11 , σ 11 ,..., µ mn , σ mn ,..., µ SK , σ SK ] µ ik µ jk
. The features can where: and are the mean values of the kth
be used to characterize the texture image.
features of class ωi and ωj respectively.
IV. SUBBAND SELECTION ALGORITHM σ ik and σ jk are the standard deviations of the kth
A feature selection algorithm that can reduce the features of class ωi and ωj respectively.
A greater G k indicates that class ω i and ω j has a better
dimensionality of feature vectors is often desirable. In the
context of wavelet transform, feature selection can be
separateness (discrimination) with the kth features.
treated as subband selection. However, the problem of
finding an optimal feature subset is NP-hard [9, 10]. In Corresponding to the weighted L 2 distance, we define
general, the fewer classes to be dealt with, the simpler the the new discriminants based on Fisher’s discriminant :
problem can be. For texture image classification, we may | µ − µ |2
divide all the candidate images into two classes. Those L 2 = ik jk
Gk
candidates similar to the sample image can be considered
σ 2 + σ jk 2
as one class, because they have similar feature vectors. ik ;
The remaining images can be treated as the other class, as We name the discriminants as “feature selection
their feature vectors can be quite different from that of functions” respectively.
the sample image. Thus, we may set a “threshold” that (3) Calculate the mean of all values computed using
can be used to select effective subbands, so that the two the ith-subband’s feature selection functions:
classes can be well separated.
The basic idea is to rank all the subbands according m 1
certain criteria so that the problem of selecting the µ is1 = ∑ s ;
G ij (j=1,2,…,m)
effective subset reduces to picking the first few subbands. i =1 m
Such subbands can be defined as effective and
discriminating ones, offering the benefit that only the In this way we obtain altogether n mean values of the
effective and discriminating features are involved in feature selection functions corresponding to n subbands :
computation during texture image classification.
We employ the feature selection functions for choosing µ is1 , … µ ijs ,…, µ in
s ;
effective Gabor wavelet subbands. First, let us define the
subband selection problem as follows. Given a texture
image that has been decomposed into n subbands, the (4) Sort the mean values ( µ s , … µ ij
s ,…, µ s ) in a
in
i1
goal is to select a set of l (l<n) subbands that can be decreasing order. Select l( l <n) subbands corresponding
effectively used to produce the minimal retrieval error. to the first l mean values.
Obviously, the number of the selected subbands should The selected l Gabor wavelet subbands can be
be kept as small as possible without affecting the final effectively used for texture image classification. In this
classification accuracy. paper, we used the statistics (mean and standard deviation)
The subband selection scheme comprises the following extracted from the selected subbands to construct texture
steps: feature vectors for classification.
(1) Suppose there are m texture images in the database. More details about the feature selection algorithm can
Compute the J-level Gabor wavelet decomposition of all be seen in [16].
images to obtain n subbands.

© 2009 ACADEMY PUBLISHER

768 JOURNAL OF COMPUTERS, VOL. 4, NO. 8, AUGUST 2009

V. EXPERIMENTAL RESULTS

A variety of experiments are performed to verify the

effectiveness of the combining scheme. Fig. 4 shows
some texture samples from the Brodatz texture database
[21].
All the texture images are transformed by wavelet
decomposition and Gabor wavelet decomposition to
extract texture feature for classification. The introduced
texture features are tested using the Brodatz texture
database [21]. In our experiments, our implementation of
the Gabor wavelet is based on Manjunath et al’s work [7,
8]. In our experiment, the feature is computed by filtering
the texture image with 6 orientations and 4 scales and Fig4. Some texture samples from the Brodatz texture database.
computing the mean and standard deviation of the output
in the frequency domain. In our experiments, we select (comprising mean and variance values), whereas the
the 12 subbands out of the total 24 subbands. Every traditional texture feature vector is 48-dimensional [7, 8].
texture is represented by a 24-dimensional feature vector Then, Support Vector Machine algorithm, which is
developed from optimal classification in the sense of
linear separation [14, 15], is used as the classifier. Kernel
functions are introduced in order to construct non-linear
decision surfaces. They map the data to some higher
dimensional feature space and construct a separation
hyper-plane in this space. Radial basis function kernel,
also called RBF:

{
K ( x, xi ) = exp − x − xi / γ 2
2
};
This is a radial basis classifier. For most pattern
recognition applications, using different parameters may
have more effects on the correct recognition rate than
using different kernel functions [14, 15].The RBF kernel
can be used in applications related to human vision
characteristics such as license plate recognition.
The SVM algorithm, using the RBF kernel functions,
employed for classification was as described in the above
section. The experiments results for various experimental
training samples are listed in Table 1. Every sample was
divided into 24 sub-images to perform the classification.
For every class, we random selecting 14 textures used for
training and the rest 10 textures for testing. Another
experiment is 18 textures used for training per class, and
the rest 6 textures were used for testing.
As the Table 1 shows, the method using feature selection
algorithm outperforms the method using all the Gabor
wavelet subbands features and visual perceptual texture
features. The feature selection algorithm successfully
selected effective Gabor wavelet subbands that can
improve correct classification rates. Our experiments
show that if we do not carried out the subband selection
procedure, although Gabor wavelet can capture more
orientation information of the sample texture than
selected effective subbands, the classification
performance is worse. We deduce that many subbands are
not effective and discriminating or redundant when used
as texture features. After the subband selection procedure,
although the dimensionality of the texture feature vector
is lower, the classification accuracy is raised. The average
accuracy produced in different ways is shown in Table 1.

© 2009 ACADEMY PUBLISHER

JOURNAL OF COMPUTERS, VOL. 4, NO. 8, AUGUST 2009 769

It can be seen that our proposed method is not only

effective but also efficient.

Table 1. Experiments results for various experimental

trainning samples.
Method A: using all the Gabor wavelet subbands
features and visual perceptual texture features.
Method B: selected Gabor wavelet subbands features
and visual perceptual texture features using feature
selection algorithm.
D43
Training Testing Average Average
samples samples Accuracy Accuracy
number number Of Of
Method A Method B
(a) (b) (c) (d)
14ⅹ102 10ⅹ102 81.6% 83.7%
18ⅹ102 6ⅹ102 89.1% 91.3%
Fig.5 Some texture images (D97, D43) in the Brodatz texture
In addition, we analyzed the cases which have low database and result of are the sub-texture-images producing by
classification accuracy and misclassifications. We divide the same texture into 4 ((a), (b), (c), (d)) sub-images.
discovered that the cause of misclassification was that
some sub-images in a same class were not perceptually
similar at all. Because some texture images in the Brodatz
texture database are anisotropic, in the phase of dividing VI. CONCLUSION AND DISCUSSION
each texture into 24 overlapping sub-texture-images, sub- In this paper, we propose a scheme for texture image
images in the same class are not similar ultimately, such classification using visual perceptual texture features and
as sub-images (b) and (c) derived from the D43 texture, Gabor wavelet features. Then, a simple feature selection
sub-images (a) and (d) from the D97 texture image procedure based on a modification of Fisher’s
(shown in Fig.5). These misclassifications of sub-texture- discriminant for texture classification. Subbands are first
images are observed dissimilar from human visual sorted according to the mean value of proposed feature
perception. The experimental results are therefore selection functions. Then the first few subbands are
reasonable if these cases are discarded. selected. Our experiments show that while the
dimensionality of feature vectors is reduced (for Gabor
wavelet: from 48 down to 24), the classification accuracy
is raised. Many experiments have tested the effectiveness
of the proposed scheme.
Furthermore, more experiments based on different image
database will be tested in the future.

ACKNOWLEDGMENT
This work is supported by the National Natural Science
Foundation of China and Chinese Academic of Sciences
joint fund on astronomy under project No. 10778619.

D97 REFERENCES
[1] Long F., Zhang H.J., and Feng D.D. (2003)
"Fundamentals of content-based image retrieval" in
Multimedia Information Retrieval and Management-
Technological Fundamentals and Applications, D. Feng,
(a) (b) (c) (d) W.C. Siu, and H.J. Zhang (Eds.), Springer, 2003.
[2] Howarth, P., Ruger, S.: Evaluation of texture features for
content-based image. retrieval. In: Proceedings of the
International Conference on Image and Video Retrieval,
Springer-Verlag (2004) 326–324
[3] H. Tamura, S. Mori, and T. Yamawaki, "Texture features
corresponding to visual perception," IEEE Trans. On
Systems, Man, and Cybernetics, vol. Smc-8, No. 6, June
1978.
[4] F. Liu, and R. W. Picard, "Periodicity, directionality, and
randomness: Wold features for image modeling and

© 2009 ACADEMY PUBLISHER

770
retrieval," IEEE Trans. on Pattern Analysis and Machine
Learning, Vol. 18, No. 7,July 1996.
[5] R. W. Picard, T. Kabir, and F. Liu, "Real-time recognition
with the entire Brodatz texture database,"Proc. IEEE Int.
Conf. on Computer Vision and Pattern Recognition, pp.
638-639, New York, June .1993.
[6] I. Daubechies, "The wavelet transform, time-frequency
localization and signal analysis," IEEE Trans. on
Information Theory,Vol.36,pp.961-1005,1990
[7] B. S. Manjunath, and W. Y. Ma, "Texture features for
browsing and retrieval of image data," IEEE Trans. on
Pattern Analysis and Machine Intelligence, Vol. 18, No. 8,
pp. 837-842, Aug. 1996.
[8] Manjunath, B. Wu, P. Newsam, S. and Shin, H. “A Texture
Descriptor for Browsing and Similarity Retrieval.” Journal
of Signal Processing: Image Communication, 16(1), pp.
3343, September 2000.
[9] Abhir H. Bhalerao and Nasir M. Rajpoot.” Discriminant
Feature Selection for Texture Classification”.In proc.
British Machine Vision Conference ,sep.2003
[10] O. Boz. Feature Subset Selection by Using Sorted Feature
Relevance. In Proc. Intl. Conf. on Machine Learning and
Applications, June 2002.
[11] Simona E. Grigorescu, Nicolai Petkov, and Peter Kruizinga
“Comparison of Texture Features Based on Gabor Filters”
IEEE Transactions on image processing, Vol. 11, no. 10,
OCTOBER 2002
[12] J. Smith and S.-F. Chang, “Transform features for texture
classification and discrimination in large image database”.
IEEE Intl. Conf. on Image Proc., 1994.
[13] Y. Ma and B. Manjunath, “Texture features of wavelet
transform features for texture image annotation”, IEEE
Intl. Conf. on Image Proc., 1995.
[14] Vapnik V N. Statistical Learning Theory [M]. New York:
John Wiley & Sons Inc, 1998.
[15] XueGong Zhang. Introduction to statistical learning theory
and support vector machines[J]. Acta Automatica
[16] Dong, Junyu; Jian, Muwei; Gao, Dawei; Wang, Shengke.
"Reducing the dimensionality of feature vectors for texture
image retrieval based on wavelet decomposition".
Proceedings - SNPD 2007: Eighth ACIS International
Conference on Software Engineering, Artificial
Intelligence, Networking, and Parallel/Distributed
Computing, v 1, 2007, p758-763..
[17] Muwei, Jian, Dong, Junyu; Gao, Dawei; Liang, Zuojuan
"New texture features based on wavelet transform
coinciding with human visual perception". Proceedings-
SNPD 2007: Eighth ACIS International Conference on
Software Engineering, Artificial Intelligence, Networking,
and Parallel/Distributed Computing, v 1, 2007, p 369-
373.
© 2009 ACADEMY PUBLISHER
JOURNAL OF COMPUTERS, VOL. 4, NO. 8, AUGUST 2009
[18] Muwei, Jian, Junyu Dong, “New Perceptual Texture
Features Based on Wavelet Transform”. International
Journal of Computer & Information Science (IJCIS).
Volume 9, Number 1, pp 11
[19] Georgeson, M. A., “Spatial Fourier Analysis and Human
Vision,” Chapter 2, in Tutorial Essays in Psychology, A
Guide to Recent Advances, N. S. Sutherland (ed.), vol 2,
Lawrence Earlbaum Associates, Hillsdale, N.J., 1979.
[20] Battiato, S. Gallo, G. Nicotra, S. Perceptive visual texture
classification and retrieval. 12th International Conference
on Image Analysis and Processing pp. 524- 529. Sept.
2003
[21] P. Brodatz, Textures: A Photographic Album for Artists&
Designers, Dover, New York, 1966.
[22] Georgeson, M. A., “Spatial Fourier Analysis and Human
Vision,” Chapter 2, in Tutorial Essays in Psychology, A
Guide to Recent Advances, N. S. Sutherland (ed.), vol 2,
Lawrence Earlbaum Associates, Hillsdale, N.J., 1979.
[23] Devalois, R.L., D.G. Albrecht, and L.G. Thorell, “Spatial -
frequency selectivity of cells in acaque visual cortex,”
Vision Research, 22, pp. 545-559, 1982.
[24] Caelli, T., Visual Perception, Pergamon Press, 1981.
Muwei Jian is a lecturer in the
School of Space Science and Physics,
Shandong University at Weihai, China.
Mr. Jian holds 3 granted patents and
has published more than 20 journal and
international conference papers
(indexed by EI and ISTP). His research
interests include image processing,
wavelet analysis, multimedia analysis
and computer vision. He has investigated issues in content-
based image/multimedia retrieval, 3D texture synthesis, editing
and synthesis of 3D surface texture, texture classification and
image fusion etc.
Haoyan GUO is a lecturer in the School of Computer Science
and Technology, Harbin Institute of Technology (Weihai),
Weihai, China.
Lei Liu is a lecturer in the Institute of Physics and
Electronic Engineering, TaiShan University, Taian, China.